A broadband ASE source configured within the telecommunications windows of optical fibers would have application as a test instrument for the communications industries (e.g., telecommunications, cable television, etc.). In general, the wavelength band over which a fiber-optic communication system works is governed by the availability of optical amplifiers for that band. It thus becomes necessary to test equipment over the same wavelengths where amplifiers exist. In the 1500 nm window of optical fibers, erbium-doped fiber amplifiers (EDFA) have proven to be adequate for fabrication of reasonably priced ASE sources. However, sources in the 1300 nm window are a different matter. Currently, semiconductor optical amplifier (SOA)-based sources are available but they are costly and very unstable as a result of the strong pumping needed to obtain the necessary high powers. Praseodymium-doped fibers have also become available, but unfortunately they are limited to covering the 1290 nm to 1315 nm spectral range, no where near the full 1280 nm to 1345 nm available in the optical fiber. Beyond test equipment, there are emerging applications in communications where broadband ASE sources can be used as transmitters themselves, such as an optical code division multiple access (O-CDMA).
Other applications have also recently been developed whereby a high-power, low-coherence, broadband source can be used in imaging. In particular, optical coherence tomography (OCT) uses the low coherence as a gating mechanism to image various tissues in the body by interfering a beam from a reference arm with one reflected from the sample tissues. Because the process is based on interference, it is sensitive to amplitude fluctuations in the source. In addition, because various tissues in the body absorb and reflect at different wavelengths, the ability to provide sources at any needed wavelength becomes a distinct advantage. Currently, only the SOA-based and EDFA-based ASE sources are commercially available for OCT. Unfortunately, these sources are also relatively expensive.
As a further example, an ASE source would have application in fiber gyroscopes. These devices work by interfering beams in two arms of an optical-fiber based interferometer. Due to the Doppler effect, light traveling parallel to the direction of rotation is accelerated and light traveling anti-parallel is slowed. Because the device is interferometric, it is extremely sensitive and requires very stable sources. Unfortunately, because it also needs to be rotated, the device needs to be compact. Current SOA sources are too bulky and unstable to be utilized in fiber gyroscopes.
Presently, there are several devices which can provide high-power, broadband, low-coherence radiation coupled into optical fibers. One class of devices is based on semiconductors and can be divided into two groups. The first group is composed of super luminescent light emitting diodes (LEDs) and the other group is based on semiconductor optical amplifiers (SOAs) that are run without any input signal, and thus generate amplified spontaneous emission (ASE). The second class of devices are based on optical fibers doped with rare-earth ions. These fibers are normally used as in-line optical amplifiers, but when used without signal inputs can generate broadband ASE.
There are a number of techniques in the art today for fabricating optical channel waveguides. These include ion-exchange in glass substrates, ion indiffusion or proton exchange in LiNbO.sub.3 substrates, pattern definition by laser ablation, photolithography of spun polymer films, and epitaxial growth and selective etching of compound semiconductor and doped crystalline films. In general, the goal of each of these channel fabrication techniques has been to produce waveguides which support a single guided mode of propagation. Also, a disadvantage of today's techniques is that they cannot be used with a significant number of useful optical materials, such as laser crystals. Further, these approaches are all limited to similar materials for use in defining the core and cladding. For example, ion implantation used in glass integrated optical structures, ion diffusion used in doped lithium niobate waveguides, vapor deposition used in sputtered, doped glass structures, or epitaxial growth used in doped crystalline devices.
In view of the above, a need exists in the industry for an optical waveguide and radiation emitting device employing the same which allows the core and cladding materials to comprise dissimilar structural and/or chemical materials, and which can be more cost efficiently produced, for example, for the various applications discussed above.